University of Southern California

Terrago Technologies, Inc.

Cyrus Shahabi

Ugur Demiryurek

Craig A Knoblock

Ching-Chien Chen

Dipsy Kapoor

Roger Zimmermann

Kun Fu

Shu-Yuen Didi Yao

Farnoush Banaei-Kashani

Bei Pan

Yan Liu

Jaffar Khoshgozaran

Dingxiong Deng

Rose Yu

Houtan Shirani-Mehr

Rolfe Schmidt

Afsin Akdogan

Yao-Yi Chiang

Linhong Zhu

A method for traffic prediction of a road network includes receiving past traffic information corresponding to multiple locations on the road network. The method further includes determining, by a processor and based on the past traffic information, temporal characteristics of the past traffic information corresponding to changes of characteristics over time and spatial characteristics of the past traffic information corresponding to interactions between locations on the road network. The method further includes predicting predicted traffic information corresponding to a later time based on the determined temporal and spatial characteristics of the past traffic information. The method further includes receiving detected additional traffic information corresponding to the later time. The method further includes updating the temporal characteristics of the traffic information and the spatial characteristics of the traffic information based on the predicted traffic information and the detected additional traffic information.

Latent space model for road networks to predict time-varying traffic

Systems and techniques for enhancing accuracy of traffic prediction include a system of one or more computers operable to receive a request relating to traffic prediction, compare a first prediction error for a first (moving average) traffic prediction model with a second prediction error for a second (historical average) traffic prediction model, calculated using a historical data set selected from previously recorded traffic data in accordance with a day and time associated with the request, select use of the first model or the second model based on the comparison of prediction errors, and provide an output for use in traffic prediction, wherein the output comes from applying the first traffic prediction model when the first prediction error is less than the second prediction error, and the output comes from applying the second traffic prediction model when the first prediction error is not less than the second prediction error.

Traffic prediction using real-world transportation data

The present disclosure relates to a short-lived throwaway index structure for generating an index from scratch in a short period of time rather than updating an index with every location change of moving objects. Rapid index construction results from the generation of Voronoi diagrams in parallel using multiple cloud servers simultaneously.

Throwaway spatial index structure for dynamic point data

Real-time high-fidelity spatiotemporal data on transportation networks can be used to learn about traffic behavior at different times and locations, potentially resulting in major savings in time and fuel. Real-world data collected from transportation networks can be used to incorporate the data's intrinsic behavior into a time-series mining technique to enhance its accuracy for traffic prediction. For example, the spatiotemporal behaviors of rush hours and events can be used to perform a more accurate prediction of both short-term and long-term average speed on road-segments, even in the presence of infrequent events (e.g., accidents). Taking historical rush-hour behavior into account can improve the accuracy of traditional predictors by up to 67% and 78% in short-term and long-term predictions, respectively. Moreover, the impact of an accident can be incorporated to improve the prediction accuracy by up to 91%.

The class of k Nearest Neighbor (k NN) queries in spatial networks has been studied in the literature. Existing approaches for k NN search in spatial networks assume that the weight of each edge in the spatial network is constant. However, real-world edge-weights are time-dependent and vary significantly in short durations, hence invalidating the existing solutions. The problem of k NN search in time-dependent spatial networks, where the weight of each edge is a function of time, is addressed herein. Two indexing schemes (Tight Network Index and Loose Network Index) are proposed to minimize the number of candidate nearest neighbor objects and reduce the invocation of the expensive fastest-path computation in time-dependent spatial networks. We demonstrate the efficiency of our proposed solution via experimental evaluations with real-world data-sets, including a variety of large spatial networks with real traffic-data.

Efficient K-nearest neighbor search in time-dependent spatial networks

A method for processing geospatial datasets corresponding to geospatial objects, the method having the steps of extracting geospatial attributes from the geospatial datasets, locating extracted geospatial attributes corresponding to a particular geospatial object at a particular point in time, and generating output indicative of the particular geospatial object at the particular point in time utilizing the located geospatial attributes.

Processing time-based geospatial data

Methods for locating a feature on geospatial imagery and systems for performing those methods are disclosed. An accuracy level of each of a plurality of geospatial vector datasets available in a database can be determined. Each of the plurality of geospatial vector datasets corresponds to the same spatial region as the geospatial imagery. The geospatial vector dataset having the highest accuracy level may be selected. When the selected geospatial vector dataset and the geospatial imagery are misaligned, the selected geospatial vector dataset is aligned to the geospatial imagery. The location of the feature on the geospatial imagery is then determined based on the selected geospatial vector dataset and outputted via a display device.

Precisely locating features on geospatial imagery

A method, computer program, and system for linking content to individual image features are provided. A section of an image is identified. A plurality of features associated with the section of the image is determined. Each of the plurality of features corresponds to at least one position within the section of the image. Content associated with the plurality of features is retrieved from a content repository. The content is linked to the plurality of features based on at least one rule. The content is then presented.

Systems and methods for linking content to individual image features

With real-world spatial networks the edge travel-times are time-dependent, where the arrival-time to an edge determines the actual travel-time on the edge. To speed up the path computation, exact and approximate techniques for computation of the fastest path in time-dependent spatial networks are presented. An exact fastest path computation technique based on a time-dependent A* search can significantly improve the computation time and storage complexity of existing approaches. Moreover, for applications with which approximate fastest path is acceptable, the approximate fastest path computation technique can improve the computation time by an order of magnitude while maintaining high accuracy (e.g., with only 7% increase in travel-time of the computed path on average). With experiments using real data-sets (including a variety of large spatial networks with real traffic data) the efficacy of the disclosed techniques for online fastest path computation is demonstrated.

Hierarchical and exact fastest path computation in time-dependent spatial networks

Document relevance is determined with respect to a region of interest (ROI). A set of location references may be associated with a set of documents. The system selects location references associated with an ROI and then selects documents corresponding to the selected location references. The selected documents can be reported or processed further. A document-location reference index can be accessed when the present system is 'online' and processing a request for documents relevant to an ROI. The document-location reference index may be generated and updated while the present system is 'offline' and not processing a request for documents. The resulting relevant documents may be provided to a user in response to a document search associated with the ROI or along with an advertisement associated with the ROI.

Dynamically linking relevant documents to regions of interest

Automatic conflation systems and techniques which provide vector-imagery conflation and map-imagery conflation. Vector-imagery conflation is an efficient approach that exploits knowledge from multiple data sources to identify a set of accurate control points. Vector-imagery conflation provides automatic and accurate alignment of various vector datasets and imagery, and is appropriate for GIS applications, for example, requiring alignment of vector data and imagery over large geographical regions. Map-imagery conflation utilizes common vector datasets as 'glue' to automatically integrate street maps with imagery. This approach provides automatic, accurate, and intelligent images that combine the visual appeal and accuracy of imagery with the detailed attribution information often contained in such diverse maps. Both conflation approaches are applicable for GIS applications requiring, for example, alignment of vector data, raster maps, and imagery. If desired, the conflated data generated by such systems may be retrieved on-demand.

System and method for fusing vector data with imagery

Systems and techniques are described for blind evaluation of nearest neighbor queries. Locations of multiple users in an original space are received. The locations in the original space are encoded into encoded locations in a transformed space. A relative proximity of the encoded locations in the transformed space is maintained after the encoding. Multiple keys corresponding to the multiple users are generated. Each key enables a reverse transformation of an encoded user location in the transformed space to an original user location in the original space. The multiple keys are provided to the corresponding multiple users, and the encoded locations in the transformed space are provided to a device. An order of computations required to reverse transform the encoded locations in the transformed space to the locations in the original space in the absence of a key is greater than a computational threshold.

Blind evaluation of nearest neighbor queries wherein locations of users are transformed into a transformed space using a plurality of keys

Provided are methods, computer programs and systems that optimize database queries using a wavelet transform of the query. Also provided are methods and systems for optimal disk placement for wavelet data.

Methods for fast progressive evaluation of polynomial range-sum queries on real-time datacubes

Systems and techniques for transmitting data stream to a client include transmitting a data segment from one of a plurality of nodes of a continuous media server to a client according to a scheduler on the node. A system includes a plurality of data processing devices, each data processing device coupled with at least one storage device. Each data processing device includes a scheduler to schedule transmission of the data segment to a client in sequence with other data segments, and a module to transmit the data segment to the client.

Growing community of inventors

Irvine, CA, United States of America

Cyrus Shahabi